How We Review Health Content
In one line
Eight steps sit between a question and a published article. One pass looks only for overstatement. Another looks only for missing safety information.
Trust should rest on a process you can inspect, not on a promise that everything here is perfect. This is the process.
Drafted from sources, checked line by line against them, and revisited when guidance moves.
Every article is drafted from published research and health authority guidance, then checked line by line against those sources. It gets one pass for overstatement and one for safety gaps. It is opened again when guidance changes, or when a reader reports a problem.
We are small, and we say so plainly. No clinician signs off on these articles.
What we have instead is a paper trail. A checking process written down. Dates on every article. Sources you can open. A record of what we got wrong.
Trust built that way can be inspected. Trust built on a promise cannot.
From scoping the question to revisiting the article two years later.
We decide what the article covers and, just as importantly, what it does not. Some things need a clinical judgment about one person. Those get marked, and the reader is sent to a health professional.
We start with guidance from major health bodies and systematic reviews, then work down to single studies. We note the study type, the group studied, the size, and the year. Those decide how much weight a finding can carry.
We write for a reader with no medical training, and define each term the first time it appears. The aim is to simplify the language without simplifying the science past the point where it is still true.
Each factual statement goes back to the study or the guidance it came from. We look for four things. Figures that have drifted. Findings stretched to a wider group than was studied. Animal results written as human ones. Estimates written as counts.
A separate pass looks for one thing only. Language that claims more than the evidence carries. Causal words on a correlation. Superlatives. Guarantees. Any hint that health outcomes sit entirely inside a person’s control.
We check the warning symptoms are there, and that readers are told when to seek help instead of managing alone. We check the cautions are there for people with relevant conditions, on relevant medicines, and for people who are pregnant.
Every article carries a reference list with links, and the dates it was published, updated, and reviewed.
We aim to open each article again at least every two years. Sooner, when major guidance changes or a reader reports a problem. Reader reports are an ordinary part of the process. Nothing about them is exceptional.
Nine of them. Naming them is what makes them easier to catch.
This process is real and it is not infallible. It has caught errors in our own published work, which is how we know it functions. It will not catch everything. That is why the Corrections Policy matters as much as this page does.
That any of this makes us right.
The “last reviewed” date on an article means editorial review. If we ever add review by a qualified clinician, we will label it medical review, and name the reviewer.
Especially about safety information that should be there and is not.
If something here contradicts current guidance, misreports a study, or leaves out safety information that matters, we want to know. Our Corrections Policy says what happens next.
First published July 2026 · Last updated August 2026 · Last reviewed August 2026
The rest of the record: Medical Disclaimer · Editorial Policy · Corrections Policy · Terms and Conditions · Privacy Policy · Cookie Policy · Contact
